Skip to content
LLMs
Qwen logo

Qwen: Qwen3 VL 235B A22B Thinking

by Qwen

Qwen: Qwen3 VL 235B A22B Thinking is a large language model from Qwen. It costs $0.400 per million input tokens and $4.00 per million output tokens. Its context window is 131K tokens.

ReasoningTool callingStructured outputOpen weightsimage input
Input / 1M tokens
$0.400
Output / 1M tokens
$4.00
Cached input / 1M
-

Not supported

Context window
131K

tokens

Who serves it cheapest

2 hosts serve Qwen3 VL 235B A22B Thinking. Same weights, same API - the price difference is pure margin and routing.

Providers serving Qwen: Qwen3 VL 235B A22B Thinking, cheapest first
ProviderInput / 1MOutput / 1MContextThroughputUptime 24h
AlibabaCheapest$0.400$4.00131K-91.4%
Novitabf16$0.980$3.95131K-99.6%

The spread between Alibaba and Novita is 1.3× for identical weights. Quantization and context limits differ, so check both columns before switching.

About Qwen3 VL 235B A22B Thinking

Qwen3-VL-235B-A22B Thinking is a multimodal model that unifies strong text generation with visual understanding across images and video. The Thinking model is optimized for multimodal reasoning in STEM and math....

Specifications

Qwen: Qwen3 VL 235B A22B Thinking specifications
Model IDqwen/qwen3-vl-235b-a22b-thinking
ProviderQwen
Context window131K tokens
Max output33K tokens
Input modalitiestext, image
Output modalitiestext
Knowledge cutoff2025-03-31
Open weightsYes - Qwen/Qwen3-VL-235B-A22B-Thinking
ReleasedSeptember 23, 2025

Cheaper alternatives

Models that cost less than Qwen3 VL 235B A22B Thinking while keeping at least half its context window and every input modality it supports.

Frequently asked

How much does Qwen: Qwen3 VL 235B A22B Thinking cost?

$0.400 per million input tokens and $4.00 per million output tokens.

What is the context window of Qwen: Qwen3 VL 235B A22B Thinking?

131K tokens, with up to 33K tokens of output per request.

Which provider serves Qwen: Qwen3 VL 235B A22B Thinking cheapest?

Alibaba at $0.400 per million input tokens - 1.3× cheaper than Novita, the most expensive of the 2 hosts serving it.

Confirm against the source: Qwen official pricing.